Geometric Design 4.1 GEOMETRIC CROSS SECTION

Transcription

1 CHAPTER 4 Geometric Design Geometric design for transportation facilities includes the design of geometric cross sections, horizontal alignment, vertical alignment, intersections, and various design details. These basic elements are common to all linear facilities, such as roadways, railways, and airport runways and taxiways. Although the details of design standards vary with the mode and the class of facility, most of the issues involved in geometric design are similar for all modes. In all cases, the goals of geometric design are to maximize the comfort, safety, and economy of facilities, while minimizing their environmental impacts. This chapter focuses on the fundamentals of geometric design, and presents standards and examples from different modes. The order of presentation of material in this chapter is to consider geometric cross sections first, then vertical alignment, horizontal alignment, superelevation, intersections, and various design details. For purposes of exposition, the order of the topics is not very important. In a typical design project, on the other hand, there is a definite order of tasks, in which the establishment of a tentative horizontal centerline usually precedes establishment of vertical alignment. This is because the elevation of the existing ground along the centerline is an important consideration in establishing the vertical alignment. The process of designing the vertical alignment begins with plotting a profile of the existing terrain, and a tentative horizontal centerline must already be established in order to do this. 4.1 GEOMETRIC CROSS SECTION The primary consideration in the design of geometric cross sections for highways, runways, and taxiways is drainage. Details vary depending on the type of facility and agency. 1,2 63

2 64 CHAPTER 4: Geometric Design Highway cross sections consist of traveled way, shoulders (or parking lanes), and drainage channels. Shoulders are intended primarily as a safety feature. They provide for accommodation of stopped vehicles, emergency use, and lateral support of the pavement. Shoulders may be either paved or unpaved. Drainage channels may consist of ditches (usually grassed swales) or of paved shoulders with berms or curbs and gutters. Figures 4.1 to 4.4 show various types of geometric cross sections. For so-called normal crown sections (that is, no superelevation or banking of the roadway, as in a horizontal curve) the traveled way slopes down from the centerline or other profile grade line; roadway cross-slopes may either be constant or varying. Where crossslopes vary, the cross section may consist of either a parabolic section or one in which Ditch Shoulder Traveled way Shoulder Ditch Foreslope Backslope FIGURE 4.1 Two-lane highway cross section, with ditches. Shoulder Traveled way Shoulder FIGURE 4.2 Two-lane highway cross section, curbed. Shoulders Traveled way Traveled way Ditch Median Ditch FIGURE 4.3 Divided highway cross section, depressed median, with ditches.

3 4.1 Geometric Cross Section 65 Traveled way Shoulders Median Traveled way FIGURE 4.4 Divided highway cross section, raised median, curbed. each lane has a constant cross-slope, but those of the outer lanes are greater than those of the inner lanes. For high-type roadways (such as freeways), cross-slopes are normally 1.5 to 2.0 percent. Shoulders or parking lanes slope away from the centerline at 2 to 6 percent. As a general rule, superelevated sections will be constructed in a single plane (including shoulders) if the rate of superelevation exceeds the normal crossslope of the shoulder. Where ditches are used, foreslopes should normally be 1:4 or flatter to prevent damage to vehicles or injury to occupants when traversed. Backslopes (cut or fill slopes) are discussed in more detail in Chapter 5. These are normally on the order of 1:2 or 1:1.5 except in rock cuts, where they may sometimes be even vertical. Use of 1:3 or flatter ditch backslopes facilitates the use of motorized equipment for maintenance, however; also, composite slopes (that is, two different slopes at different distances from the roadway) are sometimes used. Standard lane widths are normally 3.6 m (12 ft.), although narrower lanes are common on older roadways, and may still be provided in cases where the standard lane width is not economical. Shoulders or parking lanes for heavily traveled roads are normally 2.4 to 3.6 m (8 to 12 ft.) in width; narrower shoulders are sometimes used on lightly traveled roads. Runway and taxiway cross sections are similar to those for highways, except that cross-slopes are limited to 1 percent, with slopes of 1.5 percent on turf-covered areas immediately outside the runway or taxiway. 3 Railway cross sections are as shown in Figure 4.5. The distance between the rails, as shown in the figure, is what is known as standard gage in North America. Other Rail m Rail Ballast Tie Subgrade FIGURE 4.5 Railroad track geometric cross section.

4 66 CHAPTER 4: Geometric Design gages are sometimes used. In the case of railroad track, the shape of the cross section is not intended to provide drainage, since the ballast, which is permeable material, serves this purpose VERTICAL ALIGNMENT The vertical alignment of a transportation facility consists of tangent grades (straight lines in the vertical plane) and vertical curves. Vertical alignment is documented by the profile. To repeat what was said in Chapter 3, the profile is a graph that has elevation as its vertical axis and distance, measured in stations along the centerline or other horizontal reference line of the facility, as its horizontal axis Tangent Grades Tangent grades are designated according to their slopes or grades. Maximum grades vary, depending on the type of facility, and usually do not constitute an absolute standard. The effect of a steep grade is to slow down the heavier vehicles (which typically have the lowest power/weight ratios) and increase operating costs. Furthermore, the extent to which any vehicle (with a given power/weight ratio) is slowed depends on both the steepness and length of the grade. The effect of the slowing of the heavier vehicles depends on the situation, and is often more a matter of traffic analysis than simple geometric design. Chapter 10, for instance, discusses the effects of slow-moving vehicles on highway capacity and level of service. As a result, the maximum grade for a given facility is a matter of judgment, with the tradeoffs usually being cost of construction versus speed. In the case of railroads, on the other hand, the tradeoff is an economic one, involving travel time, construction cost, and minimum power/weight ratios for trains on various grades. Table 4.1 gives the maximum grades recommended for various classes of roadway by AASHTO. It should be understood that considerably steeper grades can be TABLE 4.1 Recommended standards for maximum grades, percent Type of terrain Freeways Rural highways Urban highways Level Rolling Mountainous Source: From A Policy on Geometric Design of Highways and Streets. Copyright 1994 by the American Association of State Highway and Transportation Officials, Washington, DC. Used by permission.

5 4.2 Vertical Alignment / lb per % of grade lb FIGURE 4.6 Force diagram for calculating grade resistance. negotiated by passenger cars. In some urban areas, grades of minor streets may be as steep as 25 percent. In the case of railways, older freight lines were built with grades of up to 2 percent. Modern practice is to limit grades against the load to 0.5 percent where possible, and grades against empty cars to 1.5 to 2.0 percent. The reason is that the rolling resistance of rail vehicles is very low compared with the grade resistance. Consequently, it is economical to build very flat grades and use low power/weight ratios for trains. Figure 4.6 illustrates the calculation of grade resistance. As can be seen, the resistance of the grade is approximately 20 lb/ton per 1 percent of grade. Rolling resistance, meanwhile, is around 3 lb/ton for level, tangent track in warm climates, where modern roller bearings are used. With this disproportion between rolling resistance and grade resistance, it makes sense to underpower trains and construct very level track, or else route lines so as to gain elevation gradually. The calculation of the distance the line needs to traverse in order to gain a certain amount of elevation, given a maximum grade, is known in railway practice as development. Rail lines used for passenger traffic only, particularly urban rail transit systems, often use much steeper grades. Grades of up to 10 percent can be negotiated. Maximum grades for light rail systems are reported to be in the range of 6 to 9 percent; those for conventional heavy rail rapid transit systems are in the range of 3 to 4 percent. In the case of urban rail systems, lighter equipment, higher power/weight ratios to provide rapid acceleration, and numerous vertical controls (station elevations, streets to be passed over or under, etc.) account for the decision to allow steeper maximum grades. In the case of runways used by aircraft with approach speeds of 121 knots or more, the maximum grade at any point is limited to 1.5 percent, and grades exceeding 0.8 percent are permitted only on the inner two quarters of the runway. Minimum grades are sometimes specified for highways. These are normally intended to provide for drainage on curbed facilities. On such facilities, the grade of the edge of shoulder is normally the same as the profile grade, so that the centerline must have some minimum grade. For curbed roadways, AASHTO recommends a minimum grade of 0.50 percent, unless pavements are constructed with a high degree of quality control, in which case 0.30 percent is acceptable.

6 68 CHAPTER 4: Geometric Design g 1 BVC x Elevation = y Point T P.I. L L/2 EVC g 2 FIGURE 4.7 Symmetrical vertical curve Vertical Curves Vertical tangents with different grades are joined by vertical curves such as the one shown in Figure 4.7. Vertical curves are normally parabolas centered about the point of intersection (P.I.) of the vertical tangents they join. Vertical curves are thus of the form y y 0 g 1 x rx2 2 where y elevation of a point on the curve y 0 elevation of the beginning of the vertical curve (BVC) g 1 grade just prior to the curve x horizontal distance from the BVC to the point on the curve r rate of change of grade The rate of change of grade, in turn, is given by (4.1) r g 2 g 1 (4.2) L where g 2 is the grade just beyond the end of the vertical curve (EVC) and L is the length of the curve. Also, vertical curves are sometimes described by K, the reciprocal of r. K is the distance in meters required to achieve a 1 percent change in grade. Vertical curves are classified as sags where g 2 g 1 and crests otherwise. Note that r (and hence the term rx 2 2) will be positive for sags and negative for crests. Also note that vertical distances in the vertical curve formulas are the product of grade times a horizontal distance. In consistent units, if vertical distances are to be in meters, horizontal distances should also be in meters, and grades should be dimensionless ratios. In many cases, however, it is more convenient to represent grades in percent and horizontal distance in stations. This produces the correct result, because the grade is multiplied by 100 and the horizontal distance divided by 100, and the two factors of 100 cancel. It is very important not to mix the two methods, however. If grades are in percent, horizontal distances must be in stations; likewise, if grades are dimensionless ratios, horizontal distances must be in meters.

7 4.2 Vertical Alignment 69 The parabola is selected as the vertical curve so that the rate of change of grade, which is the second derivative of the curve, will be constant with distance. Note that the first derivative is the grade itself, and since the rate of change of grade is constant, the grade of any point in the vertical curve is a linear function of the distance from the BVC to the point. That is, g dy (4.3) dx g 1 rx The quantity rx 2 2 is the distance from the tangent to the curve and is known as the offset. If x is always measured from the BVC, the offset given by rx 2 2 will be measured from the g 1 tangent. To determine offsets from the g 2 tangent, x should be measured backward from the EVC. Since the curve is symmetrical about its center, the offsets from the g 1 and g 2 tangents, respectively, are also symmetrical about the center of the curve, which occurs at the station of its P.I. Other properties of the vertical curve may be used to sketch it. For instance, at its center, the curve passes halfway between the P.I. and a chord joining the BVC and EVC. At the quarter points, it passes one quarter of the way between the tangents and the chord. Normal drafting practice is to show the P.I. by means of a triangular symbol, as in Figure 4.7, although the extended vertical tangents shown in the figure are often omitted. The BVC and EVC are shown by means of circular symbols. The P.I., BVC, and EVC are identified by notes. The stations of the BVC and EVC are given in notes, as are the station and elevation of the P.I., the two tangent grades, and the length of the vertical curve. Elevations on vertical curves are easily calculated by means of a calculator, computer, or spreadsheet program. One traditional way of representing them is in the form of a table, such as Table 4.2. The table represents a 300 m sag vertical curve between TABLE 4.2 Tabular form for profile calculations in vertical curve Tangent Profile Station Grade elevation Offset elevation % BVC P.I EVC %

8 70 CHAPTER 4: Geometric Design a 1.0% grade and a 6.0% grade. The first column gives the station. The second column gives the intersecting grades and the locations of the BVC, P.I., and EVC. The third column gives the elevation of each point on the tangent grades, calculated as BVC elevation plus g 1 x for the first tangent grade and P.I. elevation plus g 2 (x L 2) for the second. The fourth column gives the offset, calculated as rx 2 2, with x measured from either the BVC or EVC as appropriate: since offsets are symmetrical about the P.I., however, they need be calculated only from the BVC to the P.I. The last column gives the curve elevation, which is the tangent elevation plus the offset. It should be noted that curve elevations can also be calculated by using only offsets from the g 1 tangent, and that in many cases it may be more convenient to use only one tangent. EXAMPLE PROBLEM 4.1 A 2.5% grade is connected to a 1.0% grade by means of a 180 m vertical curve. The P.I. station is and the P.I. elevation is m above sea level. What are the station and elevation of the lowest point on the vertical curve? Rate of change of grade: r g 2 g 1 L Station of the low point: At low point, g 0 g g 1 rx 0 or x g 1 r 1.0% ( 2.5%) 1.8 sta 1.944%/sta a 2.5 b sta Station of BVC (100 00) (0 90) Station of low point (99 10) (1 29) Elevation of BVC: y m ( 0.9 sta)( 2.5% ) m Elevation of low point: rx 2 y y 0 g 1 x 2 (1.944%/sta)(1.29 sta) m ( 2.5% )(1.29 sta) m Design standards for vertical curves establish their minimum lengths for specific circumstances. For highways, minimum length of vertical curve may be based on sight distance, on comfort standards involving vertical acceleration, or appearance criteria. For passenger railways (especially urban rail transit systems), minimum vertical curve lengths will often be based on vertical acceleration standards; for freight railways, much more stringent standards may be maintained to avoid undue stress on couplings

9 4.2 Vertical Alignment 71 Height of eye Line of sight S Object Height of Object FIGURE 4.8 Stopping sight distance diagram for crest vertical curve. in sag vertical curves. For airport runways and taxiways, minimum vertical curve lengths are based on sight distance. In most cases, sight distance or appearance standards will govern for highways. The equations used to calculate minimum lengths of vertical curves based on sight distance depend on whether the sight distance is greater than or less than the vertical curve length. For crest vertical curves, the minimum length depends on the sight distance, the height of the driver s eye, and the height of the object to be seen over the crest of the curve, as illustrated in Figure 4.8. The minimum length is given by the formula (4.4) where S sight distance (from Table 3.3) L vertical curve length A absolute value of the algebraic difference in grades, in percent, g 1 g 2 h 1 height of eye h 2 height of object For stopping sight distance, the height of object is normally taken to be m. For passing sight distance, the height of object used by AASHTO is m. Height of eye is assumed to be m. Inserting these standard values for h 1 and h 2, Equation (4.4) may be reduced to for stopping sight distance and for passing sight distance. 200(1h 1 1h 2 ) 2 L min µ 2S 200(1h 1 1h 2 ) 2 L min µ L min µ AS 2 A AS S 404 A AS S 946 A when S L when S L when S L when S L when S L when S L (4.5) (4.6)

10 72 CHAPTER 4: Geometric Design 1 S m FIGURE 4.9 Stopping sight distance diagram for sag vertical curve. For sag vertical curves, stopping sight distance is based on the distance illuminated by the headlights at night. Design standards are based on an assumed headlight height of m and an upward divergence of the headlight beam of 1. This is illustrated by Figure 4.9. As in the case of crest vertical curves, the formulas for minimum length of vertical curve depend on whether the length of the curve is greater or less than the sight distance. For sag vertical curves, the formula is AS 2 AS S(tan 1 ) S L min µ S(tan 1 ) S 2S 2S A A when S L when S L (4.7) Design charts or tables are used to determine minimum length of vertical curve to provide stopping sight distance for both crest and sag vertical curves, and passing sight distance on crests. These may be found in the AASHTO Policy on Geometric Design of Highways and Streets. In some cases, sag vertical curves with a small total grade change can be sharp enough to cause discomfort without violating sight distance standards. In this case, it is necessary to establish a comfort criterion of the form r a v 2 (4.8) where r is the rate of change of grade, a is the maximum radial acceleration permitted, and y is speed. There is no general agreement as to the maximum value of radial acceleration that can be tolerated without producing discomfort. AASHTO suggests a value of 0.3 m/s 2, and suggests the standard L AV2 395 (4.9) where L length of vertical curve, m A g 2 g 1, percent V design speed, km/h Comfort standards for passenger railways are based on similar vertical accelerations.

11 4.2 Vertical Alignment 73 For high-speed main tracks for freight railways, the specifications of the American Railway Engineering Association (AREA) are much more restrictive. The rate of change of grade is limited to 0.10 percent per station for crest vertical curves and 0.05 percent per station for sags. If the train is traveling down a grade steep enough to overcome rolling resistance, the couplings between the cars will be compressed. If the center of gravity of the train passes the low point of the curve while it is still compressed, the slack is suddenly pulled out of the couplers, and the resulting jerk may break the train apart. The design standard is intended to ensure that as the train approaches the low point, cars will be on grades too flat to overcome rolling resistance, and the slack will be pulled out of the couplers gradually. For airports, rates of change of grade are limited to 0.33 percent per station on runways at airports used for high-speed aircraft. In addition, the Federal Aviation Administration (FAA) requires that for airports with no control tower, or at which the control tower is not always in operation, a 1.5 m object be visible from any point on the runway with a 1.5 m eye height. Minimum vertical curve standards for highways may also be based on appearance. This problem arises because short vertical curves tend to look like kinks when viewed from a distance. Appearance standards vary from agency to agency. Current California standards, for instance, require a minimum vertical curve length of 60 m where grade breaks are less than 2 percent or design speeds are less than 60 km/h. Where the grade break is greater than 2 percent and the design speed is greater than 60 km/h, the minimum vertical curve is given by L 2V, where L in the vertical curve length in meters and V is the design speed in km/h. EXAMPLE PROBLEM 4.2 Determine the minimum length of a crest vertical curve between a 0.5% grade and a 1.0% grade for a road with a 100-km/h design speed. The vertical curve must provide 190-m stopping sight distance and meet the California appearance criteria. Round up to the next greatest 20 m interval. Stopping sight distance criterion: Assume S L L AS 2 200(1h 1 1h 2 ) ( 1.0)4(1902 ) 200( ) m m 190 m, so S L L 2S 200(1h 1 1h 2 ) 2 A m 2(190) 200( ) ( 1.0)4 Appearance criterion: Design speed 100 km/h 60 km/h but grade break 1.5% 2%. Use 60 m. Conclusion: Sight distance criterion governs. Use 120 m vertical curve.

12 74 CHAPTER 4: Geometric Design EXAMPLE PROBLEM 4.3 Determine the minimum length of a sag vertical curve between a 0.7% grade and a 0.5% grade for a road with a 110 km/h design speed. The vertical curve must provide 220 m stopping sight distance and meet the California appearance criteria and the AASHTO comfort standard. Round up to the next greatest 20 m interval. Stopping sight distance criterion: Assume S L L 65.3 m 220 m, so S L AS S 30.5 ( 0.7)4(2202 ) 65.3 m (220) L 2S S A 2(220) m (220) 30.5 ( 0.7)4 Since L 0, no vertical curve is needed to provide stopping sight distance. Comfort criterion: L AV ( 0.7)4(1102 ) 36.8 m 395 Appearance criterion: Design speed 110 km/h 60 km/h but grade break 1.2% 2%. Use 60 m. Conclusion: Appearance criterion governs. Use 60 m vertical curve. Finally, vertical curve lengths may be limited by the need to provide clearances over or under objects such as overpasses or drainage structures. In the case of sag vertical curves passing over objects or crest vertical curves passing under them, the required clearances establish minimum lengths; in the case of crest vertical curves passing over objects or sags passing under them, the clearances establish maximum lengths. Where clearances limit vertical curve lengths, adequate sight distance should still be provided. In either case, the maximum or minimum length of the vertical curve may be determined by assuming that the clearance is barely met and calculating the length of the vertical curve passing through the critical point thus established. It is easiest to do this as illustrated by Figure In the figure, C represents the critical clearance, z the horizontal distance from the P.I. to the critical point, and y the offset between the critical point and the tangent passing through the BVC. The equation for the offset is where r, as before, is y rx2 2 (4.10) r g 2 g 1 L A L (4.11)

13 4.2 Vertical Alignment 75 BVC g 1 P.I. C y g 2 EVC z Object FIGURE 4.10 Calculation of lengths of vertical curves constrained by clearances. and x L 2 z (4.12) Substituting Equations (4.9) and (4.10) into Equation (4.8) results in y A(L 2 z)2 2L (4.13) Expansion and rearrangement of Equation (4.11) leads to the quadratic equation AL 2 (4Az 8y )L 4Az 2 0 (4.14) Solving Equation (4.14) results in two roots. The smaller of these represents a vertical curve that is tangent between the P.I. and the critical point. Discarding this solution and letting w y A to simplify the notation, the solution for the in the larger root leads to L 4w 2z 42w 2 wz as an expression for the maximum or minimum vertical curve length. (4.15) EXAMPLE PROBLEM 4.4 A vertical curve joins a 1.2% grade to a 0.8% grade. The P.I. of the vertical curve is at station and elevation m above sea level. The centerline of the roadway must clear a pipe located at station by 0.80 m. The elevation of the top of the pipe is m above sea level. What is the minimum length of the vertical curve that can be used? BVC 1.2% P.I. Elevation m m z y EVC +0.8% Elevation m Pipe

14 76 CHAPTER 4: Geometric Design Determine z: Determine y Determine w: Determine L: Check y : z (75 40) (75 00) 0.40 sta. Elevation of tangent ( 1.2)(0.4) m Elevation of roadway m y m A g 2 g 1 ( 0.8) ( 1.2 ) 2.0 w y A L 4w 2z 42w 2 wz x r A L (0.74) 2(0.4) (0.74)(0.4) 4.17 sta 417 m sta y rx2 2 (0.48)( ) Check 4.3 HORIZONTAL ALIGNMENT Horizontal alignment for linear transportation facilities such as highways and railways consists of horizontal tangents, circular curves, and possibly transition curves. In the case of highways, transition curves are not always used. Figure 4.11 illustrates horizontal alignments with and without transition curves Horizontal Tangents Horizontal tangents are described in terms of their lengths (as expressed in the stationing of the job) and their directions. Directions may be either expressed as bearings or as azimuths and are always defined in the direction of increasing station. Azimuths are expressed as angles turned clockwise from due north; bearings are expressed as angles turned either clockwise or counterclockwise from either north or south. For instance, the azimuth 280 is equivalent to the bearing north 80 west (or N80 W). Figure 4.12 illustrates azimuths and bearings.

15 4.3 Horizontal Alignment 77 With transition curves Tangent Circular curve Circular curve Transition curve Transition curve Transition curve Transition curve Tangent Without transition curves Tangent Circular curve Tangent Tangent Circular curve FIGURE 4.11 Horizontal alignments with and without transition curves. N Azimuth = 30 Bearing = N30 E Azimuth = 290 Bearing = N70 W Azimuth = 210 Bearing = S30 W FIGURE 4.12 Azimuths and bearings Circular Curves Horizontal curves are normally circular. Figure 4.13 illustrates several of their important features. Horizontal curves are described by radius (R), central angle ( ) (which is equal to the deflection angle between the tangents), length (L), semitangent distance (T),

16 78 CHAPTER 4: Geometric Design P.I. TC T L C E M CT R /2 FIGURE 4.13 Horizontal curve features. middle ordinate (M), external distance (E), and chord (C). The curve begins at the tangent-to-curve point (TC) and ends at the curve-to-tangent point (CT). In the past, severity of curvature was sometimes expressed in degree of curvature. Although obsolete in the metric system, degree of curvature may still be encountered in some situations. Degree of curvature may be defined in two ways. The arc definition is the angle subtended by a 100 ft arc. The chord definition is the angle subtended by a 100 ft chord. The relationship between radius (in feet) and degree of curvature (arc definition) is D 36,000 (4.16) 2pR R where D degree of curvature and R radius of curvature, in feet. The length of a circular curve is given by L 2pR (4.17) 360 R rad where is the central angle of the curve; rad refers to measured in radians. The semitangent T of a circular curve is given by T R tan 2 (4.18) The middle ordinate M is given by The external distance E is given by M R R cos a 2 b E R cos( 2) R (4.19) (4.20)

17 4.3 Horizontal Alignment 79 Point 1 TC d 1 c 1 d 2 c 2 Point 2 FIGURE 4.14 Layout of circular curve. The chord C is given by C 2R sin a 2 b (4.21) Circular curves are usually laid out in the field by occupying the tangent-to-curve point TC with a transit and then establishing successive points by turning deflection angles and measuring chords, as shown in Figure The deflection angle in radians d x to a point on the curve at a distance x from the TC is given by d x a x (4.22) 2R b rad The chord c x to this point is given by c x 2R sin d x (4.23) Table 4.3 gives deflection angles and chords at 20 m intervals for a 500 m radius curve with a deflection angle of 15 and a TC at station Design standards for horizontal curves establish their minimum radii and, in some cases, their minimum lengths. Minimum radius of horizontal curve is most commonly TABLE 4.3 Circular curve layout table Radius 500 m Deflection angle Station x, m Radians Degrees Chord, m TC CT

18 80 CHAPTER 4: Geometric Design Sight distance s m Line of sight Eye R Obstruction R Object FIGURE 4.15 Stopping sight distance on horizontal curves. established by the relationship between design speed, maximum rate of superelevation, and curve radius, which is discussed in Section 4.4. In other cases, minimum radii or curve lengths for highways may be established by the need to provide stopping sight distance or by appearance standards. Figure 4.15 illustrates the relationship between curve radius, stopping sight distance, and the setback distance to obstructions to vision. The relationship between the radius of curvature R, the setback distance m, and the sight distance s is given by m R c 1 cos a 28.65s R bd (4.24) and s R (4.25) c cos 1 a R m bd R where the angles in the formulas are measured in degrees. Since these formulas are hard to solve for R, design charts or tables are normally used to find the minimum radius of curvature that will provide stopping sight distance. Minimum lengths or radii of horizontal curves may also be based on appearance criteria. Where deflection angles are small, a short horizontal curve may give the appearance of a kink. To prevent this, minimum horizontal curve lengths may be prescribed for curves with small deflection angles Transition Curves Transition curves are used to connect tangents to circular curves. Several forms of curve have been used for this purpose. The most logical choice from a theoretical standpoint, and the only one discussed here, is the clothoid spiral, for which the radius of curvature varies as the inverse of the distance along the curve from its beginning.

19 4.3 Horizontal Alignment 81 Spirals are used both for esthetic reasons and because they provide a rational superelevation transition. In the case of railways, such rational superelevation transitions are virtually necessary for reasons of vehicle dynamics. The use of spirals in conjunction with superelevation transitions is discussed in Section 4.4. In the case of highways, spirals are used primarily for esthetic purposes. They are most appropriate for roadways with relatively high design standards, where largeradius curves are used. Under these circumstances, drivers can often see a considerable way ahead on the roadway, and can detect the difference between the more smooth, flowing lines provided by the transition curves and the more abrupt alignment that results in their absence. For roadways with lower design standards, some recent research reports indicate that use of spirals may increase accident rates. It is surmised that the reason is that drivers have a harder time judging the severity of curves where spirals are used. 6 Figure 4.16 shows how spirals connect circular curves to tangents, and illustrates the nomenclature of spirals. Critical points in moving through the curve are the tangent to spiral point (TS), the spiral to curve point (SC), the curve to spiral point (CS), and the spiral to tangent point (ST). The effect of using the spirals is to shift the circular portion of the curve inwards, so that it no longer fits to the original tangents. It now fits to what are called offset tangents, which are shifted in by a distance p, measured perpendicular from the original tangents. The distance from the TS to any point on the curve is L; if L is measured for the entire distance from the TS to the SC, it is referred to as L s. Likewise, the angle between the tangent and a line tangent to the spiral at any distance L is referred to as the spiral angle u; if the spiral angle is measured at the SC, it is referred to as u s. The distance along the extended tangent from the TS to a point opposite that at which the circular curve is tangent to the offset tangents is referred to as k, and the distance from this point to the P.I. as T. The length of the circular portion of the curve is L c, and the coordinates of any point on the spiral, measured relative to the tangent, are X and Y. TS k X s T P.I. L s SC Y s p L c R c s CS s L s ST FIGURE 4.16 Spiral transition curve nomenclature.

20 82 CHAPTER 4: Geometric Design Since the spiral is defined as the curve such that the reciprocal of the radius varies linearly from zero at the TS to 1 R c at the SC, or The spiral angle u is given by In particular, 1 R a 1 R c ba L L s b (4.26) RL R c L s A 2 (4.27) u L 2R (4.28) u s L s 2R c (4.29) The coordinates X and Y of any point on the spiral may be expressed as functions of L: and X L Y L3 6A 2 L5 40A 4 L9 3,456A 8 L7 336A L ,240A (4.30) (4.31) Other measurements of interest are p Y s R c (1 cos u s ) (4.32) k X s R c sin u s (4.33) T (R c p) tan a 2 b (4.34) and L c R c ( rad 2u s ) R c rad L s (4.35) Tables exist which allow X and Y to be calculated without directly evaluating Equations (4.28) and (4.29); however, since these equations are readily evaluated by computers or programmable calculators, such tables are of less importance than formerly, and will not be discussed further here. Spirals are laid out in the field in a manner similar to that for circular curves. In this case the TS is occupied by the transit, and successive points along the spiral are established by turning deflection angles and measuring chords. For a point on a spiral

21 4.4 Superelevation 83 TABLE 4.4 Spiral curve layout table Spiral angle U Deflection angle d Station L X Y Radians Degrees Radians Degrees c TS SC whose coordinates have previously been calculated, the deflection angle d is d tan 1 a Y X b (4.36) and the chord c is c 2X 2 + Y 2 (4.37) Table 4.4 gives X, Y, u, d, and c for 20 m intervals along an 80 m spiral that connects a tangent with a 500 m radius circular curve, with the TS at station In order to determine the stations of the critical points on the curves, it is necessary to remember that stationing runs along the curves and not the tangents. The station of the TS may be calculated measuring from the ST point of the last curve to the P.I. (which is actually not on the centerline of the facility), establishing a temporary station at the P.I., and subtracting T k from this temporary station. The station of the SC may be found by adding the length of the spiral to the station of the TS; that of the CS is found by adding the length of the circular portion of the curve, as given by Equation (4.35), to that of the SC; and the station of the ST is found by adding the length of the spiral to that of the CS. 4.4 SUPERELEVATION The purpose of superelevation or banking of curves is to counteract the centripetal acceleration produced as a vehicle rounds a curve. The term itself comes from railroad practice, where the top of the rail is the profile grade. In curves, the profile grade line follows the lower rail, and the upper rail is said to be superelevated. Since most railways are built to a standard gage, the superelevations are given as the difference in elevation between the upper and lower rail. In the case of highways, somewhat more complicated modifications of the cross section are required, and, because widths vary, superelevation is expressed as a slope. Consider the force diagram in Figure If the vehicle is traveling around a curve with a radius R at a constant speed v, there will be a radial acceleration toward

22 84 CHAPTER 4: Geometric Design W W v 2 g R BANKS F 2 F 1 N 1 N 2 W F W v 2 g R W sin N = W cos FIGURE 4.17 Force diagram for superelevation. the center of the curve (toward the left in the diagram) of v 2 R, which will be opposed by a force of (W g)(v 2 R). Other forces acting on the vehicle are its weight W and the forces exerted against the wheels by the roadway surface. These forces are represented by two components: the normal forces N 1 and N 2 and the lateral forces F 1 and F 2. For highway vehicles F 1 and F 2 are friction forces, so F 1 mn 1 and F 2 mn 2 (4.38) where m is the coefficient of friction between the tires and the roadway. Summing forces parallel to the roadway gives cos u a W g bav2 R b F W sin u (4.39) Defining a factor f so that f K F N (4.40) and noting that N W cos u sin u a w g bav2 R b (4.41)

23 4.4 Superelevation 85 Equation (4.37) may be rewritten as cos u a W g bav2 R b W sin u fw cos u f sin u a w g bav2 R b (4.42) Dividing by W cos u leads to v 2 v2 tan u f f tan u a gr gr b (4.43) But tan u is the cross-slope of the roadway, which is the same as the superelevation rate e. Consequently, v 2 v2 e f ef a (4.44) gr gr b or v 2 (1 ef) e f (4.45) gr The term ef is small compared to one, and may be omitted, so the relationship can be simplified to Also, v 2 gr f e R v 2 g( f e) A commonly used mixed-unit version of Equation (4.45) is (4.46) (4.47) R V 2 127( f e) (4.47a) where V is in km/h and R is in meters. Alternatively, e v2 gr f or e V 2 127R f (4.48) Values of f recommended by AASHTO are conservative relative to the actual friction factor between the tires and the roadway under most conditions, and vary with design speed. These are given in Table 4.5. Maximum rates of superelevation are limited by the need to prevent slow-moving vehicles from sliding to the inside of the curve and, in urban areas, by the need to keep parking lanes relatively level and to keep the difference in slope between the roadway and any streets or driveways that intersect it within reasonable bounds. AASHTO recommends that maximum superelevation rates be limited to 12 percent for rural

24 86 CHAPTER 4: Geometric Design TABLE 4.5 Values of side friction recommended by AASHTO Design speed, km/h Maximum side friction factor Source: From A Policy on Geometric Design of Highways and Streets. Copyright 1994 by the American Association of State Highway and Transportation Officials, Washington, DC. Used by permission. roadways; 8 percent for rural roadways for which snow or ice are likely to be present; and 6 percent or 4 percent for urban streets. In addition, there is a tradeoff between the maximum rate of superelevation and the minimum curve radius permitted at any design speed. AASHTO recommends the minimum curve radii shown in Table 4.6. For the higher design speeds, the superelevation rates for these minimum curve radii, as calculated by Equation (4.48), are less than the maximum superelevation rates given above. Consequently, the maximum superelevation rates really apply only to fairly low design speeds. EXAMPLE PROBLEM 4.5 What is the minimum radius of curvature allowable for a roadway with a 100 km/h design speed, assuming that the maximum allowable superelevation rate is 0.12? Compare this with the minimum curve radius recommended by AASHTO. What is the actual maximum superelevation rate allowable under AASHTO recommended standards for a 100 km/h design speed, if the value of f is the maximum allowed by AASHTO for this speed? Round the answer down to the nearest whole percent. Minimum radius of curvature for 100 km/h design speed: R Minimum radius recommended by AASHTO is 490 m. Actual maximum superelevation rate for AASHTO recommended standards for 100 km/h is e V 2 127( f e) ( ) 328 m V2 127R f (490) Rounding, e max %

25 4.4 Superelevation 87 TABLE 4.6 Recommended minimum radius of curvature Design speed, km/h Manimum curve raidus, m Source: From A Policy on Geometric Design of Highways and Streets. Copyright 1994 by the American Association of State Highway and Transportation Officials, Washington, DC. Used by permission. Superelevation transitions involve modification of the roadway cross section from normal crown to full superelevation, at which point the entire roadway width has a cross-slope of e. The manner in which this transition is accomplished is expressed by a superelevation diagram, which is a graph of superelevation (cross-slope) versus distance measured in stations. As an alternative, the diagram may show the difference in elevation between the profile grade and the edge versus distance. Figure 4.18 is an example superelevation diagram, showing the transition from normal crown with 2 percent cross-slopes to 6 percent superelevation for a roadway with a spiral transition curve. Figure 4.19 is the alternative form of the diagram, assuming a two-lane highway with 3.6 m lanes. Figure 4.20 presents an interpretation of the superelevation diagram, showing the appearance of the cross section at intervals 6% Tangent runoff TS Superelevation runoff L SC Rate of superelevation 4% 2% 0 2% 4% 6% Station FIGURE 4.18 Superelevation diagram, showing rate of superelevation.

26 88 CHAPTER 4: Geometric Design Difference in elevation, profile grade minus edge, mm Tangent runoff TS Superelevation runoff L Station SC FIGURE 4.19 Superelevation diagram, showing difference in elevation. 6% TS L SC Rate of superelevation 4% 2% 0 2% 4% 6% Station 2% 2% 0% 2% 2% 4% 6% FIGURE 4.20 Superelevation diagram, showing roadway cross sections. through the transition. As shown in Figure 4.18, the superelevation transition is normally linear; that is, the rate of rotation of the cross section is constant with respect to distance through the transition. The distance marked L, which runs from the point at which the outside half of the roadway (that is, the half on the outside of the curve) is at zero cross-slope to the P.I. at full superelevation (or from the tangent-to-spiral point TS

27 4.4 Superelevation 89 to the spiral-to-curve point SC), is called superelevation runoff. The distance from the point at which the outside half of the roadway first begins to rotate to the TS is referred to as tangent runoff. As previously mentioned, spiral transition curves are often used in conjunction with superelevation transitions, and normally coincide with the superelevation runoff. The clothoid spiral provides for a rational superelevation transition because 1 R varies linearly with L. From Equation (4.44), f e r V2 (4.49) R Consequently, for constant V, f e is proportional to 1 R. If a linear superelevation transition is used, e is a linear function of L, where L is the distance from the beginning of the transition to any point. Meanwhile, 1 R is also proportional to L; therefore, f must also be a linear function of L. Consequently, side friction increases linearly through a clothoid spiral superelevation transition if it is traversed at constant speed. In the case of railways, use of transition curves in conjunction with superelevation transitions is virtually necessary. Railway superelevation is normally designed so that f 0 if the curve is traversed at design speed. The reason for this is that nonzero side forces may lead to binding of the flanges of the wheels against the rails. More importantly, lateral forces tend over time to force the track out of its proper alignment and gage. Consequently, the usual practice is to design for zero side friction. Note that the clothoid spiral is the only superelevation transition curve that can meet this criterion fully. If f is set to zero, e is proportional to 1 R; since 1 R varies linearly with L, so can e. For highways, superelevation is usually applied by rotating the cross section about the profile grade line. This will be the centerline in the case of two-lane highways, undivided multilane highways, and multilane divided highways with paved medians. For multilane divided highways with wide medians, the inside edge of traveled way is often used as the axis of rotation. For freeway ramps, the axis of rotation will normally be the edge of traveled way closer to the freeway; however, it may be shifted to avoid sags in the edge grade of the ramp. For railways, the axis of rotation is the top of the lower rail. The length of the superelevation runoff L is determined by either vehicle dynamics or appearance criteria. Where spiral transition curves are used, the superelevation transition will normally coincide with the spiral, as shown in Figure In the case of railways, the minimum spiral length may be established by the need to limit the rate of increase of centripetal acceleration in the spiral, which leads to the following formula: L V3 RC (4.50) where L minimum length of spiral, m V speed, km/h R curve radius, m C rate of increase of centripetal acceleration, m/s 3

28 90 CHAPTER 4: Geometric Design e SC g D L TS Edge Centerline FIGURE 4.21 Difference in grade resulting from superelevation transition. A value of C of 1 has usually been used in railroad practice. The formula has also sometimes been used for highways, with values of C ranging from 1 to 3. More commonly, superelevation transition lengths for highways are based on appearance or comfort criteria. One such criterion is a rule that the difference in longitudinal slope (grade) between the centerline and edge of traveled way of a two-lane highway should not exceed Figure 4.21 illustrates the application of this rule. L is measured from the TS to the SC, as in the superelevation diagram. At the TS the difference in elevation between the centerline and edge is zero. At the SC it is the superelevation rate e times the distance D from the centerline to the edge. Thus the difference in grade between the centerline and the edge is g De (4.51) L Since the criterion that the difference in grade not exceed implies that De L (4.52) L is given by L 200De (4.53) L is normally rounded up to some convenient length, such as a multiple of 20 m. EXAMPLE PROBLEM 4.6 A two-lane highway (3.6 m lanes) with a design speed of 100 km/h has a 400 m radius horizontal curve connecting tangents with bearings of N75E and S78E. Determine the superelevation rate, the length of spiral if the difference in grade between the centerline and edge of traveled way is limited to 1 200, and the stations of the TS, SC, CS, and ST, given that the temporary station of the P.I. is The length of the spiral should be rounded up to the next highest 20 m interval.

29 4.4 Superelevation 91 Determine superelevation rate: e V2 127R f (400) Determine length of superelevation transition and spiral: L s 200De 200(3.6)(0.08) 57.6 m Determine spiral angle and coordinates of SC point: u s L s 60 (0.075) rad 2R c 2(400) Round to 60 m A 1L s R c 1(60)(400) X s L s Y s L3 s 6A 2 L5 s 40A m 603 6(154.9) (154.9) ,240(154.9) m Determine p, k,, T, and L c : p Y s R c (1 cos u s ) [1 cos(0.075)] m k X s R c sin u s sin(0.075 ) m (90 75 ) (90 78 ) rad T (R c p) tan a 2 L9 s 3,456A (154.9) ,456(154.9) 8 L7 s 336A L 11 s 6 42,240A 10 b ( ) tan a 27 2 b m L c R c rad L s 400(0.471) m Determine stations of critical points: TS station P.I. station (T k) (150 00) [(0 96.1) (0 30.0)] SC station TS station L s ( ) (0 60)

30 92 CHAPTER 4: Geometric Design CS station SC station L c ( ) (1 28.4) ST station CS Station L s ( ) (0 60) COORDINATION OF HORIZONTAL AND VERTICAL ALIGNMENT Transportation facilities such as highways and railways are three-dimensional objects. Although many aspects of their design can be determined by considering horizontal and vertical alignment separately from one another, it is important to understand the relationship between them. Proper coordination of horizontal and vertical alignment is important for reasons related to the esthetics, economics, and safety of the facility. As a general rule, horizontal curvature and grades should be kept in balance. That is, the designer should avoid both the provision of minimal curvature at the expense of long, steep grades and the provision of level vertical alignment at the expense of excessive horizontal curvature. Where there is both horizontal and vertical curvature, it is normally best from an esthetic standpoint to provide the impression of a single three-dimensional curve in both the horizontal and vertical planes. This means that horizontal and vertical curves should normally coincide. In some cases, however, safety considerations may suggest that horizontal curves be extended beyond vertical curves in order to avoid hiding the beginning of the horizontal curve from drivers approaching it. This is especially important where horizontal curves coincide with rather sharp crest vertical curves. In addition, use of long, relatively gentle curves and short tangents will normally produce a more flowing line than will use of long tangents and short, sharp curves. This is especially true when the deflection angle of horizontal tangents or the difference between grades is small. This effect is illustrated by Figure At the same time, too much curvature may also pose problems. On two-lane highways, for instance, it is also important to provide an adequate number of sections with passing sight distance, and these sections need to be of adequate length to prevent drivers from becoming impatient when following slow-moving vehicles. Finally, on intersection approaches, highway alignments should provide adequate sight distance and should be as flat and straight as possible. In addition to these rules, there are a number of guidelines related to the positioning of curves (either horizontal or vertical) relative to one another. These are based on esthetics and in the case of horizontal curves, on the necessity of providing adequate superelevation transitions. Figure 4.23 illustrates several special curve combinations.